DETECTION OF BROKEN ROTOR FAULTS IN 3-PHASE INDUCTION MOTOR USING MCSA BY DIFFERENT WAVELET TRANSFORMS TECHNIQUES

Abstract

Induction motors being robust, efficient, low cost and reliable in operation and are widely used as electrical drives in industries. So, safe operation of induction machines is a major problem. Among the different mechanical faults experienced by induction motor, rotor faults contribute to 5%-10%. According to location of fault the rotor faults can be categorized into cracked bar, End ring, bar ring joint, melted bars, air gap eccentricity etc... The motor current signature analysis (MCSA) is commonly employed for detection of these faults due to less cost, reliable and non-intrusive in nature. Conventional FFT based analysis of MCSA has many drawbacks like poor resolution, noise problems etc. Therefore, in this paper, the MCSA using different wavelet transforms like Discrete Wavelet Transforms (DWT), Stationary Wavelet Transforms (SWT), and Wavelet Packet decomposition (WPD) have been proposed and compared. The proposed topologies have been tested by experimental setup interfacing with MATLAB programming.